EvoColony: A Hybrid Approach to Search-Based Mutation Test Suite Reduction Using Genetic Algorithm and Ant Colony Optimization

Abstract

The increasing complexity of software systems requires robust and efficient test suites to ensure software quality. In this context, mutation testing emerges as an invaluable method for evaluating a test suite’s the fault detection capability. Traditional approaches to test case generation and evaluation are often inadequate, particularly when applied to mutation testing, which aims to evaluate the quality of a test suite by introducing minor changes or mutations to the code. As software projects increase in scale, there is greater computational cost of employing exhaustive mutation testing techniques, leading to a need for more efficient approaches. Incorporating metaheuristics into the realm of mutation testing offers a synergistic advantage in optimizing test suites for better fault detection. Especially, combining test suite reduction methods with mutation testing produces a more computationally efficient approach compared to more exhaustive ones. This study presents a novel approach, called EvoColony, which combines intelligent search-based algorithms, specifically genetic algorithms and ant colony optimization, to reduce test cases and enhance the effectiveness of the test suit for mutation testing. Integrating both metaheuristic techniques, the research aims to optimize existing test suites, and to improve mutant detection with fewer test cases, thus improving the overall testing quality. The results of experiments conducted were compared with traditional methods, demonstrating the superior effectiveness and efficiency of the proposed hybrid approach. The findings show a significant advancement in test case reduction when using the hybrid algorithm with mutation testing methodologies, and thus ensure the quality of test suites. © 2024, Ismail Saritas. All rights reserved.